I have searched the questions regarding notation of "such that" in this forum but only encountered discussions on the use of : vs. | and less common alternatives.
In some cases I find myself in need of using two "such that" formulations in a nested fashion in set builder notation, consider the following example:
We consider some set of points $\textbf{x} \in \mathcal{X} \subset \mathbb{R}^n$ and some vector of functions $\textbf{r} : \mathbb{R}^n \to \mathbb{R}$.
Now let the subset $\mathcal{F} = \{ \textbf{x} \in \mathcal{X} : \textbf{r}(\textbf{x}) \leq \textbf{0} \}$ be called the set of feasible points (where the inequality is understood to hold pointwise), then its complement could be denoted as $\{ \textbf{x} \in \mathcal{X} : \exists i : r_i(\textbf{x}) > 0 \}$.
While I suppose this notation is valid, it is not very intuitive as it relies on right associativity of the existential quantifier which (at least IMHO) is not very satisfying.
Instead I would actually consider a mixture of notations such as $$\{ \textbf{x} \in \mathcal{X} : \exists i | r_i(\textbf{x}) > 0 \}$$ easier to understand.
- Is there a downside to this mixed notation that I am unaware of?
- Is there some common practice for nested such that notation?